What is fancy PCA?

What is fancy PCA?

PCA Color Augmentation (also called Fancy PCA) alters the intensities of the RGB channels along the natural variations of the images, denoted by the principal components of the pixel colors (Bargoti & Underwood, 2016). It performs Principal Components Analysis on the color channels, thus, given the name Fancy PCA.

What is label preserving transformation?

Note the term label preserving transformations refers to the fact that if image x is an element of class y then φ(x) is also an element of class y.

How to implement fancy PCA image augmentation in Python?

How to implement ‘Fancy PCA’ image augmentation in Python from the paper “Imagenet Classification With Deep Convolutional Neural Networks” Code: here The second form of data augmentation consists of altering the intensities of the RGB channels in training images.

Where did the term PCA color augmentation come from?

The term PCA Color Augmentation refers to a type of data augmentation technique first mentioned in the paper titled ImageNet Classification with Deep Convolutional Neural Networks. This paper is famous for introducing the AlexNet convolutional neural network architecture, which won the 2012 ImageNet Large Scale Visual Recognition Competition.

How does fancy PCA improve ImageNet accuracy?

Implementing Fancy PCA augmentation into my training appeared to increase the accuracy of the model from ~83% on the evaluation set to ~85%. In comparison, the authors of the paper noted ~1% accuracy improvement for the ImageNet challenge due to Fancy PCA.

How are eigenvalues used in fancy PCA training?

Fancy PCA uses eigenvalues and eigenvectors from Principal Component Analysis to augment an image while maintaining the features and detail of the image. Implementing Fancy PCA augmentation into my training appeared to increase the accuracy of the model from ~83% on the evaluation set to ~85%.